Questions: Genome Assembly

Question 1

What does an N50 value of 50 kb mean for a genome assembly?

• A. The assembly contains exactly 50 contigs
• B. 50% of the reads are longer than 50 kb
• C. Half of the total assembled sequence is contained in contigs of 50 kb or longer
• D. The average contig length is 50 kb

Answer: (C)

N50 is calculated by sorting contigs from longest to shortest and summing their lengths until the cumulative sum reaches 50% of the total assembly size. The length of the contig that crosses the 50% threshold is the N50. An N50 of 50 kb means that contigs of 50 kb or longer collectively account for at least half the assembly. It is a measure of contiguity, not count or average.

Question 2

Repetitive sequences longer than the sequencing read length do not pose any special challenge for genome assembly.

• T. True
• F. False

Answer: False

Repeats longer than the read length are the primary challenge in genome assembly. When a read falls entirely within a repeat, it is identical (or nearly so) to reads from other copies of the same repeat elsewhere in the genome. The assembler cannot determine which genomic location the read came from, leading to collapsed repeats (multiple copies assembled as one), misjoins, or breaks in the assembly at repeat boundaries. This is why long-read technologies, which can span entire repeat elements, dramatically improve assembly contiguity.

Question 3

Explain the difference between de novo assembly and reference-guided assembly, and when you would choose each approach.

Model answer: De novo assembly reconstructs the genome from reads alone, without using any existing reference sequence. It is necessary when no closely related reference exists (novel organisms, highly divergent strains) and when you want to detect structural variants, novel sequences, or rearrangements that a reference would miss. Reference-guided assembly maps reads to an existing reference genome and calls variants relative to it. It is faster, less computationally demanding, and appropriate when a high-quality reference from a closely related organism is available and the goal is to identify variants rather than discover novel genomic content.

In practice, many projects use both: reference-guided assembly for variant calling and a de novo assembly to capture sequences absent from the reference. The human genome reference (GRCh38) is the backbone of most human genomics, but de novo assembly of individual genomes has revealed substantial structural variation missed by reference-guided approaches.